Genesys updates contact center AI agent, orchestration tools

Genesys previewed an updated contact center AI agent to be released later this year, along with tools that orchestrate teams of agents -- including those from other vendors -- as open interoperability standards evolve to handle such tasks.

Products shown at the Genesys Xperience 2025 user conference in Nashville include role-based copilots to assist frontline contact center agents, managers, business leaders and admins; additional language support; and detailed agent performance dashboards.

The upcoming contact center AI tools come on the heels of a hot summer for Genesys. The company received massive investments from Salesforce and ServiceNow, and also released a development platform, AI Studio, for its users to create their own agents.

Genesys also plans to support the rapidly emerging Agent2Agent Protocol (A2A) and Model Context Protocol standards that manage AI agents and the information they access. Users can control agent behavior with AI Guides, a no-code development tool that sets up guardrails for large language model-based agents and includes testable safety controls.

The A2A standard is the basis for an expanded Genesys-ServiceNow partnership, which gives joint users further ability to coordinate AI agents. At Xperience, Genesys planned a live demonstration of agents from the two companies working together across the two platforms. But the company isn't stopping there; it intends to support and automate CX processes bidirectionally, connecting with CRM, service, billing and any other apps that support an aspect of customer experience.

"We're also talking with Salesforce, Adobe and others," said Mike Szilagyi, senior vice president and general manager of product management at Genesys. "Just about every application stack out there is going to have an ability to create a virtual agent on their platform and expose it to whatever tools exist inside their enterprise environment."

Agentic AI tools based on LLMs are more accurate than those of previous generations of AI technology because they can derive context from what people say, Szilagyi said.

In the contact center, AI agents, for example, can catch when people misspeak a number and correct themselves, something that would have tripped up older AI tools.

That helps in both automated conversations and those where AI is helping humans on the phone perform "slot collection," or getting a particular piece of information to move a process forward. Examples might include eliciting the last four digits of a social security number to authenticate a user or dates for booking a hotel stay.

"It's made the conversations with humans much more human-like and less error-prone," Szilagyi said.

As enterprise cloud customers roll out agentic AI, security risks are emerging. Users should start with low-risk use cases for the technology, according to Ian Jacobs, an analyst at Opus Research. Risks they take into consideration should not only include those that incur direct costs -- such as fraud -- but also indirect costs such as damage to reputation or actions that could trigger the attention of regulators.

That said, the upside of agentic AI tools like Genesys previewed is substantial, Jacobs said. A company can transform from being a price leader in its market to being a service leader, available 24 hours a day, with multilingual support, serving many channels, including phone, chat and popular messaging apps such as WhatsApp.

"Shifting a brand's whole reputation -- and a brand's promise -- is a potential outcome of this kind of technology," Jacobs said. "People are really nervous [about agentic AI], but if you're not looking at how big a change this could possibly make for your business, you're going to err on the side of being too conservative."

Don Fluckinger is a senior news writer for Informa TechTarget. He covers customer experience, digital experience management and end-user computing. Got a tip? Email him.